1. Field of the Invention
The present invention relates to an image processing apparatus, an image pickup apparatus, a computer, an image processing method and an application thereof such as a computer readable non-transitory medium, and more particularly, to technologies concerned with compression processing of image data.
2. Description of the Related Art
In the field of image pickup apparatuses such as cameras, a digital photographed image is acquired by using an imaging element (such as a CCD and a CMOS) provided with color filters such as RGB. Captured image data is stored and outputted in various image formats, such as RAW data that is digital data formed by using an electric signal from an imaging element, and a predetermined image format such as a JPEG
In particular, according to the JPEG format for compressing image data to reduce the amount of data, the compression rate is relatively high but degradation of image quality is relatively inconspicuous, and therefore the JPEG format is widely used in the field of image pickup apparatuses. Concerning this image compression processing such as a JPEG format, various techniques are proposed.
For example, Japanese patent application publication No. 2003-289542 discloses an image information encoding method capable of quantization based on visual characteristics. According to the method disclosed in Japanese patent application publication No. 2003-289542, different quantization parameters of QPluma and QPchroma are used in compression processing of a luminance signal and a color difference signal respectively. Although weighting of the two quantization parameters can be changed, two arrays A (QP) and B (QP) corresponding to a parameter QP meet the relation, A (QP)×B (QP)=Const. (“Const.” means a constant). A quantization matrix can be set by a user in units of pictures, but if the quantization matrix is not set, a predetermined default value is used.
In addition, Japanese patent application publication No. 2002-354266 discloses an image processing apparatus that selects a quantization coefficient corresponding to color components of an input image to reduce degradation of image quality due to image compression processing. According to this image processing apparatus disclosed in Japanese patent application publication No. 2002-354266, the ratio of data components of at least one of luminance data and color difference data is calculated and a quantization parameter is selected in accordance with the calculation result.
Further, Japanese patent application publication No. 11-168745 discloses a digital camera that records digital image pickup data as well as an array of color filters corresponding to respective pixels, as filter array data, in order to reduce a burden of data capacity in a recording medium and enable pixel interpolation desired by a user.
As the number of pixels in an imaging element is increased in recent years, the data size (amount of data) of image pickup data tends to bloat as a whole. Thus, compressing and storing image pickup data while reducing degradation of image quality is one of issues to be solved in a photographing technical field.
In addition, although a high frequency component of image data is removed by using an optical low-pass filter and the like to reduce image degradation such moire, it is possible to acquire a photographed image in which image degradation such moire is sufficiently reduced without using a low-pass filter by devising an array pattern of color filters. Thus, it is possible to acquire a photographed image in which resolution feeling is further improved without positively using an optical low-pass filter and without losing a high frequency component, depending on an array pattern of color filters. However, it is assumed that image data including such a high frequency component may not be sufficiently reduced in size by conventional compression processing, or may not sufficiently take advantage of superiority of including a high frequency.
As described above, a configuration and functions of an image pickup apparatus are also diversified such as presence or absence of an optical low-pass filter, use of color filters of various array patterns, or the like. Unfortunately, it is difficult for the conventional techniques disclosed in Japanese patent application publication Nos. 2003-289542 and 2002-354266 to perform appropriate image compression processing corresponding to the above-described diversification.
For example, the technique disclosed in Japanese patent application publication No. 2003-289542 may enable quantization corresponding to visual characteristics such as preventing a pseudo contour from occurring, but does not optimize image compression processing with respect to the presence or absence of an optical low-pass filter and change of an array pattern of color filters. In addition, Japanese patent application publication No. 2002-354266 discloses an example of a compression processing technique of selecting a quantization coefficient corresponding to color components of an input image, but the technique disclosed in Japanese patent application publication No. 2002-354266 also does not optimize image compression processing with respect to the presence or absence of an optical low-pass filter and change of an array pattern of color filters, as with the technique disclosed in Japanese patent application publication No. 2003-289542. Further, the technique disclosed in Japanese patent application publication No. 11-168745 also does not optimize image compression processing with respect to the presence or absence of an optical low-pass filter and the like.
Thus, it is difficult for the techniques disclosed in Japanese patent application publication Nos. 2003-289542, 2002-354266 and 11-168745 to compress and store image data with a high degree of efficiency while reducing degradation of image quality depending on the presence or absence of an optical low-pass filter and the like.